Re: Loops (was Re: do { quit; } else { })

On 16.04.2025 16:21, Scott Lurndal wrote:

bart <bc@freeuk.com> writes:

On 16/04/2025 06:35, Janis Papanagnou wrote:

 

>
So here, 0% of such loops have a step other than 1. THAT's what I mean
by 'unusual'.

 
Sample bias.
 

>
Scot Lurndal gave examples comparing C with BASIC/FORTRAN that used
steps other than one, to back up a claim that they routinely use 3
components. Actually they rarely do.

 
I made no such claim.  You said that for loops were not three-way
constructs - I showed that was not an accurate statement.  I made
no claims about anything being 'routine'.
 
A common construct:
 
   entity *ep = entities[0];
   for(size_t i = 0ul; i < MAX_ENTITIES; i++, ep++) {
      ep->field = do_something(args);
   }
 
in C++
 
   for(x = map.begin(); x != map.end(); ++x) {
     /* process a red-black tree element */
   }
 

There's also some other options in C++, say, STL's [functional]
'for_each()' or, in the newer standards, the variant we already
saw upthread mentioned as one hiding some "technical details".

Janis